Summaries of Articles

M any researchers have studied the impact of corporate income tax (CIT) rates on foreign direct investment (FDI) in general, and cross-border mergers and acquisitions (M&As) in particular.The consensus in the literature is that higher CIT rates decrease cross-border M&A activity.However, existing estimates of the relationship between CIT rates and cross-border M&As vary widely due to differences across studies in their chosen periods of analysis, set of countries involved, andcriticallytheir econometric modeling and estimation techniques.The lack of a reliable benchmark estimate limits the ability of policy makers to evaluate precisely the ramifications of possible CIT reforms.Against this backdrop, the main objective of this paper is to quantify the effect of CIT rates on the volume of cross-border M&A activity using econometric methods that have been developed in a parallel literature on the determinants of bilateral trade flows.We demonstrate in the process the importance of introducing these econometric refinements to the tax literature.To pursue this objective, we employ a database that includes domestic and cross-border M&As for more than 100 countries, 1995-2019, and we adapt state-of-the-art methods from the trade and FDI gravity literature, which, by analogy with physics, implies that cross-border flows ought to depend on respective country sizes (as measured by population or output) and distance (physical, cultural, and political), both bilaterally and relative to the rest of the world.We demonstrate that the choice of model specification and estimator matters for estimating the effects of CIT rates on cross-border M&As.Depending on the econometric specification, our estimated CIT effects range from being statistically indistinguishable from zero to being large, negative, and statistically significant.Three features that have become standard in the trade and FDI gravity literature stand out for estimating the effects of taxes on cross-border M&As.Specifically, these are (1) the use of the Poisson Pseudo Maximum Likelihood (PPML) estimator, (2) the inclusion of domestic (in addition to international) M&As, and (3) properly controlling for multilateral remoteness in addition to standard proxies for bilateral M&A frictions.Our preferred, most narrowly identified econometric specification implies that a 1 percentage point increase in CIT rates decreases the number of cross-border M&As by 0.8 percent relative to domestic M&As.The size of this effect is roughly intermediate